A Crystal Ball For Commuters

Researchers at the University of California at Berkeley are creating a system that may let drivers avoid traffic problems before they've even happened. The Freeway Performance Measurement System collects real-time traffic information from sensors on roads, and uses historical data to predict where congestion might develop and how drivers can best avoid it. Users can access the system via Web-enabled computers, personal digital assistants, or cell phones.

Chao Chen, a University of California at Berkeley graduate student who works on the project, says the system is better than other traffic-modeling systems because it gives drivers a concrete recommendation of the best route to take to avoidtraffic congestion.

John Wolf, a manager for the California Department of Transportation, which is backing the project, says its primary interest in the Freeway Performance Measurement System was using it to better understand the highway system. Making it available to drivers is a secondary bonus. He says that "people really hate the unpredictability" of traffic congestion more than the congestion itself, and predicts that systems like this one will help drivers cope better with their commutes by keeping them informed.

The project is slated to debut in Los Angeles this month.